Hi there Welcome to my Homepage!

Hi! I am a 3rd-year PhD. student at Peking University.

My research interests include Embodied AI, 3D/4D Reconstruction, Digital Human and other emerging areas in AI. I enjoy exploring diverse research directions and collaborating with researchers across different fields.

Feel free to reach out if you are interested in collaboration or potential opportunities.

Main News

  • 2026.04 🎉🎉 Two papers accepted to SIGGRAPH 2026 (1 conferecne, 1 journal).
  • 2026.02 🎉🎉 Two paper accepted to CVPR 2026.
  • 2025.07 🎉🎉 One paper accepted to ICCV 2025.
  • 2025.01 🎉🎉 One paper accepted to ICLR 2025.

Experience

Peking University
2023.09 - Present
Ph.D. Student, advised by Prof. Ronggang Wang
Harbin Institute of Technology
2019.9 - 2023.7
B.Eng. in Computer Science

Publications

(* equal contribution · ✉ corresponding author)

wog
ATGS: Anchored Temporal Gaussian Splatting for Long Volumetric Video Representation
Jiahao Wu*, Jie Liang*, Die Hu, Jiayu Yang, Kaiqiang Xiong, Xiaoyun Zheng, Xiang Li, Chao Wang ✉, Ronggang Wang ✉
Revealing the mechanisms of long-sequence, complex volumetric video reconstruction.
ACM TRANSACTIONS ON GRAPHICS [SIGGRAPH'2026]   [arXiv] [code] [Project page]
wog
ClipGStream: Clip-Stream Gaussian Splatting for Any Length and Any Motion Multi-View Dynamic Scene Reconstruction
Jie Liang*, Jiahao Wu*, Chao Wang, Jiayu Yang, Xiaoyun Zheng, Kaiqiang Xiong, Zhanke Wang, Jinbo Yan, FengGao, Ronggang Wang
A streaming dynamic reconstruction strategy based on fragment-wise training.
Conference on Computer Vision and Pattern Recognition (CVPR), 2026   [arXiv] [Project page]
wog
LocalDyGS: Multi-view Global Dynamic Scene Modeling through Adaptive Local Feature Decoupling.
Jiahao Wu, Rui Peng, Jianbo Jiao, Jiayu Yang, Luyang Tang, Kaiqiang Xiong, Jie Liang, Jinbo Yan, Runling Liu, Ronggang Wang
Extend dynamic reconstruction from small-scale motions to complex motions in large-scale scenes.
International Conference on Computer Vision (ICCV), 2025   [arXiv] [code] [Project page]
wog
Swift4D: Adaptive Divide-and-Conquer Gaussian Splatting for Compact and Efficient Reconstruction of Dynamic Scenes.
Jiahao Wu, Rui Peng, Zhiyan Wang, Lu Xiao, Luyang Tang, Kaiqiang Xiong, Ronggang Wang
Achieve fast dynamic reconstruction through motion–static decoupling.
International Conference on Learning Representations (ICLR), 2025   [arXiv] [code]

Services

  • Reviewer for CVPR’2026, ICLR’2026, ECCV’2026.

Talks